3 ms·
Yep and embeddings can be decoded back into meaningful text if you have the model weights and the tokenizer.
by diwank 2y ago
Yep and embeddings can be decoded back into meaningful text if you have the model weights and the tokenizer.
loading thread…